On Thursday, June 22, Online Dialogue co-founder Ton Wesseling (as speaker) and Lead Conversion Manager Ruben (as guest) hosted the Growth Marketing Summit attended in Germany. This congress has been organized for years by konversionsKRAFT, a successful CRO agency in Germany. Here are some highlights and insights.

The culture of experimentation at Netflix

Colin McFarland, director of Experimentation Platform & Core Data Science at Netflix, talked about five ingredients for the experimentation culture at Netflix.

1. Leadership buy-in

Everyone in the organization has knowledge of experimentation, including executives, resulting in the right mindset. A great example is when CEO Reed Hastings teaches Netflix employees in a video course what A/B testing is

2. Transparency around plans, expectations, gains and losses

Expectations, plans and results for major experiments are summarized in memos and transparent to everyone in the organization, ensuring different points of view.

3. Freedom & responsibility, not freedom from responsibility

Experiments are expected for most decisions - the default is to question decisions that have not been tested. Anyone may make decisions as long as they are tested.

4️. Contribution to the community

All managers and leaders are expected to contribute to ‘functional excellence. Creating forums makes it possible to share and receive feedback. It is everyone's responsibility to protect the data culture.

5️. Platform maturity

The experimentation platform plays a crucial role within Netflix's culture. Therefore, they are continuously building the platform to make Netflix and its culture successful.

When growth inhibits experimentation

Ton Wesseling, co-founder of Online Dialogue, has worked in the field for more than 20 years. He won an Experimentation Legend of the Year Award last week at Experimentation Elite in London. He talked about three things that cause experimentation to inhibit an organization's growth.

1. Structure issues: the organization and its processes
2. Science issues: data quality of experimentation
3. Substance issues: quality of ideas and test hypotheses

These problems can be solved with WOW. Three key points that must be balanced for success:

• Workflows: Validation should be ‘business as usual. Teams can continue to build features, these are then validated when they are released on the website. Thus, experimentation is in the standard operating procedure of product teams.
• Observations: Create a rigorous process for understanding/applying statistics and test results. Ensure data quality and that it is easy to use and aligned with business impact.
• Winners: Work where the opportunity for business impact is high, and don't be afraid of bigger steps. Provide good research, and evidence-based prioritization to test the right thing.

... Just a quick note: we send out a newsletter every three weeks that includes the latest blogs, team updates and, of course, news about the offerings in our academy. Click here to subscribe.

Newsletter sign up

Some other key take-aways

Daphne Tideman: Ensure a good balance of Explorers and Planners in growth teams. Explorers are good at finding new growth opportunities and driving change. Planners push the process and get more out of what already exists. Explorers, however, move too fast and Planners stay with an idea too long. Therefore, both are necessary in team. If the first experiment does not result in an improvement, keep trying and move on to something new after the returns diminish.

Simone Neeling,CRO specialist at VodafoneZiggo: Be a CAT person to help others experiment. Curious to stakeholders' goals, Analytical to analyze them and Thoughtful to help and educate stakeholders. This also means becoming friends with your IT department, since those are whole most important stakeholders for experimentation.

Marianne Stjernvall also known as the Queen of CRO, is a familiar face at CRO events. According to Marianne, there are five facets to an outstanding experimentation program:

1. Data: Ensure 95% data quality. Compare the data in your data platform with the actual data in the back-end.
2. Team: Provide multidisciplinary teams to carry out the experimentation process well.
3. Process: Focus on existing processes, and try to optimize them with experimentation. Don't go trying to completely replace processes.
4. Strategy: The strategy for experiments should fit an organization's experimentation maturity and the number and quality of experiments.
5. Buy-in: Demonstrate the value of experimentation for buy-in from leaders within the organization.

Kenda MacDonald: Funnels are not a reality, this is not how a (potential) customer behaves. Instead, focus on problems and reward activation using marketing automation. Match the different stages of the buying process with appropriate activations to hold attention. Also match content strategically throughout the customer journey to build profiles and find out what stage they are in, create multiple opt-in opportunities and build content for data.

GMS23 is worth visiting

Besides the speakers mentioned, there were some other great speakers. Since Frankfurt is easily accessible by train, this congress is definitely worth attending! Want to read more about interesting events at home and abroad? Then subscribe to our newsletter. Here you will also find the latest news, blogs and, of course, our own events and meetups.